Hi there, I’m going through the CLV basic guide on pymc-marketing. I was wondering if I should first categorise the data into a few segments then create a CLV model for each segment. I would segment based on one or more early variables such as the customers first months transactions or first transaction amount. Afterwards I would train a model for each segment. There are a number of left/right tail outcomes that Im looking at trying to model better. Hoping for any guidance on improving prediction or next steps after the quick setup code.